mCRM R Shiny user guide
The migration collision risk model (mCRM) shiny application is a
stochastic adaptation of the Band (2012) migration collision risk
worksheet. The tool can be found at
https://hidefdevo.shinyapps.io/mCRM
.
The web application is a wrapper around the
r mig_stoch_crm() function in the {stochLAB}
package
https://www.github.com/HiDef-Aerial-Surveying/stochLAB.
The web application contains default values for 70 species of migratory
birds as well as polygons that represent their approximate migratory
pathways. The mCRM tool does a couple of things: * Creates population
estimates in wind farms by sampling migratory pathways via straight
lines drawn between UK and non-UK coastlines * Runs a stochastic version
of the migratory collision risk model based on the population estimates
and user-input parameters.
The main interface
In the main interface, you’ll see a couple of buttons across the header at the top. The “hamburger” (three horizontal lines) can be used to hide the sidebar. Clicking on the version number will take you to the Github NEWS page. The “Github” logo (looks like a cat) will take you to the github repo. And if you want to report a bug, you could click on the bug, which would take you to the Issues page.
In the sidebar, you will see the four steps to follow to generate your models. When you open the application, you start on Step 1: Turbine & Wind farm features, which you will see in the main display window.
Step 1: Turbine & Wind farm features
First choose if you want to use the built-in list of windfarm
polygons, or if you want to use a custom shapefile.
Click on the “select wind farms” input box to bring up the dropdown
menu. When the wind farms are selected, click the “update windfarm list”
button.
To delete a windfarm from the list, simply click on it and hit DELETE on
your keyboard, then click the “update windfarm list” button again.
Using a custom shapefile
First, create a polygon shapefile that has a NAME field. Every polygon in your shapefile should have a UNIQUE identifier in the NAME field.
Select the dropdown option for “custom windfarms” and then click
“Browse”.
Ensure that you select all the files associated with the
shapefile
Click on the Load Shapefile button to bring up the new dropdown list with the new windfarm names.
Click on Update Windfarm List and the tabs with the windfarm data will be generated.
Wind farm parameters
After clicking the “Update windfarm list” button, tabs will appear on the screen. Each tab contains inputs for the parameters associated with the wind farms.
The inputs are filled with default values but can be updated based on specifics for the wind farms.
To update the data table, simply double click on it and it will open in editing mode. Hit CTRL + Enter to finish editing.
NOTE Although migration collision risk is only calculated assuming birds travel through a site once (not multiple times over months), monthly assignments are used to take an average of the available wind speed in the site (i.e., average activity of turbine blades). I.E., if the migratory period lies between March and June, then the average wind availability for the turbines is used in the calculation of collision risk. This value has negligible impact on the overall collision risk when compared to other elements (e.g., avoidance rates), but was incorporated to keep with conventions in the standard stochastic collision risk model.
Step 2: Species features
Clicking on the sidebar tab “Step 2: Species features” will bring you to the Species selection page. As per the windfarm selection step, click on the dropdown input box to bring up a list of species.
Select species and click on the “update species list” button. This will generate a series of tabs for each species where you can input the species parameters to use in the model. Many of the species parameters are already input, but any of these can be altered as needed.
Migratory pathway and period
Using the toggles for each species, you can select which migratory periods will be modelled. The migratory pathways are visualized in the right-hand map which shows how the pathways overlap with your chosen windfarms.
Step 3: Generate scenarios
Clicking on step 3, you can now generate population scenarios that will be fed forward into the mCRM. For every species x windfarm combination, a set of lines drawn through the migratory polygons that extend from the UK to “elsewhere” is sampled and the proportion of that sample that intersects the windfarm polygon is used to calculate a potential population estimate within the windfarm footprints.
If your species and windfarm parameters are set, then clicking the “generate scenarios” button (the green cog) will start the bootstrapping process.
Once the scenarios are generated, it is recommended that you download the worksheet. This will help with reproducibility and for debugging. It will also mean that you can bypass steps 1 and 2 and simply upload your scenarios, then run the tool.
Calculation of population estimates
Estimates of the population size are done by way of a bootstrapping exercise where lines are randomly generated through a migration corridor. Migration corridors as polygon shapefiles were generated by the British Trust for Ornithology and are built into the tool in the “all_polygons.rda” object, which can be found in the “data” folder in the Github repository.
First, a series of points spaced 1km apart were generated around England, Scotland (including Shetland), and Wales (UK points) representing one set of origin or departure points. Then a series of points were generated around Iceland, continental Europe, and in the North Atlantic randomly (to ensure birds arriving from the North Atlantic were captured) to represent another set of origin or departure points (non-UK points).
For each species in the database, a migratory corridor polygon shapefile is used to clip the UK and non-UK points.
